UTS — Universal Tracking System

The problem

UTS is HeavyTask's cross-platform tracking platform, and when I took it over the complaints were concrete: slow startup, sluggish screens, and a release process where every build to the stores was a manual, error-prone afternoon. The product worked; the engineering around it didn't scale.

My approach

I lead the platform's technical direction, coordinating a cross-functional team of 5+ developers.

• Architecture restructuring — I profiled the app before touching anything, then reorganized module loading, deferred SDK initialization off the startup path, and fixed the first-screen data waterfall so the app paints real data fast.
• Hot-path optimization — the heaviest list screens got memoized rows, display-size images, and flattened component trees, measured against a low-end Android reference device rather than a simulator.
• Release automation — I built the CI/CD pipeline with GitHub Actions and Expo EAS: tag-triggered production builds, auto-incremented build numbers, and store submission with no manual steps. Internal preview builds give QA the latest changes without waiting on store review.
• Team practices — I mentor the junior developers on the team and established the coding standards, review process, and unit-testing strategy the platform ships under.

Challenges

• Optimizing without a rewrite — the wins had to land incrementally in a production app with active users, so every restructuring step shipped behind the normal QA cycle instead of as a big-bang branch.
• Making releases boring — the pipeline's real job is removing human typing from the release path. Fifty-plus production releases later, a release is a git tag and a checklist, not an afternoon.

Outcome

App performance improved by 40% and load times dropped by 60%; the build-and-release pipeline is fully automated across iOS and Android. Private client work — the source is not public.